?DESCRIPTION (provided by applicant): Acute myeloid leukemia (AML) is an aggressive disease with poor survival and antiquated treatments poorly tolerated the elderly. Current leukemia treatment fails to kill non-dividing, leukemia stem cells, which promote relapse post-remission. For this reason, understanding the biology of these stem cells is of great interest to the field. GSK3? is a serine/threonine kinase that affects several pathways critical for cancr development. GSK3? degrades ?-catenin, a key promoter of AML stem cells, and activates NF-?B, a key protein promoting AML proliferation and survival. Preliminary studies show GSK3? is elevated and more nuclear in AML. This nuclear GSK3? more potently suppresses leukemogensis in vivo and more potently activates NF-?B signaling than cytoplasmic GSK3?. We observe that AML non-stem cells express more nuclear GSK3? than AML stem cells, and hypothesize that this increased nuclear GSK3? may shift cells from a ?-catenin-driven, self-renewing stem phenotype to a more NF-?B-driven, proliferative, non-stem phenotype. We will study the implications of GSK3? nuclear subcellular localization in AML by assessing its anti-?-catenin and pro-NF-?B effects, defining a specific mechanism of action for each effect. We will show high nuclear GSK3? specifically inhibits AML stem cell formation and maintenance, leading to less AML proliferation, colony formation, and mouse engraftment. We will show moderate levels of nuclear GSK3? promote AML progression in vivo. Finally, we will test if GSK3? levels, activity, and localization in AML stem cells or non-stem cells correlate to better overall survival and/or complete remission, predicting more nuclear GSK3? correlates with a more positive prognosis.